6-Alkoxy-5-aryl-3-pyridinecarboxamides, a New Series of

Moreover, in all cases, leptin levels were even higher than in the vehicle group, .... gel 60 F254 plates, and spots were visualized with UV light or ...
1 downloads 0 Views 2MB Size
Article pubs.acs.org/jmc

6‑Alkoxy-5-aryl-3-pyridinecarboxamides, a New Series of Bioavailable Cannabinoid Receptor Type 1 (CB1) Antagonists Including Peripherally Selective Compounds Stephan Röver,* Mirjana Andjelkovic, Agnès Bénardeau, Evelyne Chaput, Wolfgang Guba, Paul Hebeisen, Susanne Mohr, Matthias Nettekoven, Ulrike Obst, Wolfgang F. Richter, Christoph Ullmer, Pius Waldmeier, and Matthew B. Wright Pharma Research and Early Development, F. Hoffmann-La Roche AG, Grenzacherstrasse 124, Basel CH-4070, Switzerland S Supporting Information *

ABSTRACT: We identified 6-alkoxy-5-aryl-3-pyridinecarboxamides as potent CB1 receptor antagonists with high selectivity over CB2 receptors. The series was optimized to reduce lipophilicity compared to rimonabant to achieve peripherally active molecules with minimal central effects. Several compounds that showed high plasma exposures in rats were evaluated in vivo to probe the contribution of central vs peripheral CB1 agonism to metabolic improvement. Both rimonabant and 14g, a potent brain penetrant CB1 receptor antagonist, significantly reduced the rate of body weight gain. However, 14h, a molecule with markedly reduced brain exposure, had no significant effect on body weight. PK studies confirmed similarly high exposure of both 14h and 14g in the periphery but 10-fold lower exposure in the brain for 14h. On the basis of these data, which are consistent with reported effects in tissue-specific CB1 receptor KO mice, we conclude that the metabolic benefits of CB1 receptor antagonists are primarily centrally mediated as originally believed.



INTRODUCTION

The CB1 receptor is part of the endocannabinoid system that derives its name from cannabis. While the use of cannabis can be traced back thousands of years,1 the molecular targets of its pharmacology have been explored in detail only in the last few decades. Starting with the isolation of the main psychoactive constituent of Cannabis sativa, Δ9-THC by Mechoulam in 1964,2 the next decades saw the isolation and characterization of both the CB13−5 and CB2 receptors6 as well as several of their physiological ligands such as anandamide7 and 2-AG.8 The field has matured, with over 400 new publications on the cannabinoid system appearing in PubMed each year. Recent reviews have indicated that this breadth of information has uncovered as many questions as it has solved.9−12 The CB1 receptor is abundantly expressed in both peripheral and the neuronal tissues.13 Targetbased drug discovery has progressed and delivered many different series of CB1-receptor antagonists.14 These include rimonabant (SR141716A, 1), which received marketing authorization in the European Union as an antiobesity medication but was later withdrawn due to increased incidence of anxiety, depression, and suicidality, and taranabant (MK-0364, 2, Figure 1) and otenabant (CP-945,598), which entered late clinical development but were never launched. Although taranabant and otenabant were derived from different structural classes, similar © XXXX American Chemical Society

Figure 1. CB1 receptor antagonists.

increased incidence of psychiatric side effects led to termination after or during phase III clinical studies.15−18 The observed psychiatric side effects appear to preclude the use of CB1 receptor antagonists for non-life-threatening diseases or chronic treatment. However, it has become apparent that the CB1 receptor has peripheral functions, suggesting that peripherally active and selective agents may have therapeutic potential with reduced psychiatric side effects. Indeed, several groups have reported series of peripherally active CB1 receptor Received: July 16, 2013

A

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 2. Peripherally selective CB1 receptor antagonists.

Figure 3. Superposition of low energy conformers of 3 (orange) and 4 (yellow) (upper panel); structures and data for 3 and 4 (lower panel).

antagonists in the past few years,19,20 with the most common strategy to derive these molecules being the addition of polar groups onto known inhibitor backbones (Figure 2). This either increased polar surface area or conferred effective Pgp-mediated transport leading to reduced CNS exposure. Our group has pursued a relatively small chemotype that has evolved into more polar and less brain penetrating molecules. We originally explored a series of rimonabant bioisosteres that

evolved into compounds such as 3 (Figure 3).21 Such substituted pyrroles had good CB1 receptor affinity and selectivity but exhibited low metabolic stability especially in rat microsomes. Metabolite profiling revealed multiple oxidative transformations; basically all substituents on the pyrrole were efficiently oxidized in rat microsomes. Rather than trying to block all metabolically labile sites we employed a scaffold hop, replacing the electron rich pyrrole with B

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 1. Synthesis of Compoundsa

a

Reagents and conditions: (a) Br2, AcOH; (b) POCl3, reflux; (c) MeOH, reflux; (d) POCl3, quinoline reflux, then water RT; (f) arylboronic acid, Pd(dppf)Cl2·CH2Cl2, Na2CO3, toluene, water, 90 °C; (g) LiOH·H2O, THF, water, reflux; (h) R2-OH, KOH, DMSO, MW, 100 °C; (i) arylboronic acid, (Ph3P)4Pd, Na2CO3, DME, water, 85 °C; (j) various amide coupling methods.

Miyaura reaction of 7 with arylboronic acids furnished the 5aryl substituted 3-pyridine carboxylic acid esters 8, which were saponified to the acids 9. This was followed by a SNAr reaction to replace the 6-chloro substituent with alkoxides to give 6-alkoxy5-aryl-3-pyridine carboxylic acids 10. This was done on small scale with potassium hydroxide as base in DMSO in a microwave reactor. During up-scaling tests, we measured the thermal stability of potassium hydroxide/DMSO mixtures in an accelerating rate calorimeter and found exothermic decomposition above 145 °C. Consequently, for larger batches, we switched to a low temperature process that employed stirring of the above potassium hydroxide/DMSO mixtures for extended times between room temperature and 40 °C.23 Similar high yields can be obtained by using this low-temperature variant of the reaction. Conventional amide coupling methods converted acids 10a−l to the corresponding amides 14, 15, and 16. This first synthetic variant allowed installing the alkoxide and amide in the last two steps of the synthesis. Alternatively, the alkoxide addition via the SNAr reaction was done as first step with the acid 11 to give 6-alkoxy derivatives 12. These, in turn, could be converted in a second step to the amides 13 and in the last step by Suzuki−Miyaura arylation to compounds 4, 14, 15, and 16. A final synthetic variant utilized the Suzuki−Miyaura arylation of 6alkoxy derivatives 12 to afford an alternative route to 6-alkoxy-5aryl-3-pyridine carboxylic acids 10. Some of the 6-alkoxy-5-aryl-3-pyridine carboxylic acids (10m−q) were produced by an alternative synthesis that started from 3-bromo-2-chloro-5-methyl-pyridine 22 (Scheme 2) using a series of steps beginning with nucleophilic aromatic substitution of 22 with alkoxides to furnish 23a−e, followed by Suzuki−Miyaura arylations to give picolines 24a−e. These were then transformed into the corresponding carboxylic acids 10m− q by bromination, hydrolysis, and permanganate oxidation.

a six-membered electron deficient heterocycle. We hypothesized that this strategy would eliminate the need for a methyl substituent on the core and additionally reduce the propensity for oxidation of the phenyl residue by rendering it less electron rich. In addition, the reduction in overall lipophilicity should decrease the susceptibility to oxidative attack by cytochrome P450s. As an initial prototype we chose to make nicotinamide 4 because modeling suggested that a low energy conformer of 4 would superpose sufficiently well onto the energy-minimized structure of 3. Compound 4 was still very lipophilic (ClogP of 5.9) with low CB1 receptor affinity (∼40-fold less potent compared to 3), however it was 4 times more stable than 3 in rat liver microsomes. Following these encouraging results, SAR exploration based on compound 4 was initiated by varying the alkoxy and aryl residues. Because subsequent molecules showed both improved affinity and metabolic stability, we also explored variations of the core and amide residue. In the course of this work, we identified several orally active 5-aryl-6-alkoxy-3-pyridinecarboxamides, including peripherally selective compounds.



CHEMISTRY We explored several variations of the synthesis of 6-alkoxy-5-aryl3-pyridinecarboxamides and finally selected the route depicted in Scheme 1, which allowed either amide or aryl variations to be performed in the final synthetic step. All variants of this synthesis started with commercially available 6-hydroxy-nicotinic acid 5, which was brominated in acetic acid, leading to 6 that was then converted with POCl3 to 6chloro-5-bromo-nicotinic acid chloride, which was then either hydrolyzed to the acid 11 with water or transformed to the corresponding methyl ester 7 by methanolysis.22 Suzuki− C

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 2. Synthesis of Compounds 10m−qa

5-Alkoxy-4-aryl-2-pyridine carboxamides (32a,b) were prepared from commercially available 2-chloro-3-pyridinol 25 by formylation (to give 26) followed by iodination to 27 as described by Wishka et al.24 The corresponding phenolate was selectively alkylated with alkylhalogenides to produce 28a and 28b in 86% and 92% yield, respectively. Suzuki−Miyaura coupling with the Pd(dppf) catalyst was regioselective and produced 29a and 29b with yields of greater than 97%. Chemoselective reduction with zinc in acetic acid provided the intermediates 30a,b, which were oxidized to the acids 31a,b with tetrabutylammoniumpermanganate in pyridine. The final step was an amide coupling that yielded the desired 5-alkoxy-4-aryl-2pyridine carboxamides 32a,b. The synthesis of the isosteric 5-alkoxy-6-aryl-2-pyridinecarboxamides 37a,b started from commercially available 2-chloro-3fluoro-pyridine 33. Suzuki−Miyaura coupling to produce 34 in 65% yield was followed by nucleophilic aromatic substitution with appropriate alkoxides and oxidation with hydrogen peroxide in acetic acid to yield the N-oxides 35a,b. The N-oxides could be converted to the nitriles 36a,b by treatment with dimethylcarbamic chloride and trimethylsilyl cyanide in acetonitrile. Two-step hydrolysis of the nitriles followed by amide coupling afforded the desired 5-alkoxy-6-aryl-2-pyridinecarboxamides 37a,b. Synthesis of pyrazines (Scheme 5) was realized from commercially available 3,5-dibromo-2-pyrazinamine 38, which could be arylated with good regioselectivity (80% yield) to give the 3-arylisomer 39. Classic diazotization-bromination chemistry furnished the 2,5-dibromo-3-aryl-pyrazine 40. The nucleophilic aromatic substitution proceeded to 41 as the major isomer and was followed by palladium catalyzed carbonylation to provide the pyrazine carboxylate 42. 5-Alkoxy-6-aryl-pyrazine-2-carboxamide 44 was obtained from 42 by saponification and coupling with (1R,2R)-2-amino-cyclohexanol. Compound 51, the pyridazine analogue of 14g, was derived by multistep synthesis from 3,6-dichloro- pyridazine 45. Nucleophilic aromatic substitution to give the monoalkoxylated product 46 was followed by iodination (to 47) and Suzuki−Miyaura coupling to provide the 3-chloro-6-alkoxy-5-aryl-pyridazine 48. Palladium-catalyzed carbonylation, saponification and coupling with (1R,2R)-2-amino-cyclohexanol provided the 6-alkoxy-5aryl-pyridazine-3-carboxamide 51.

a Reagents and conditions: (a) R2-OH, NaH, DMF; (b) arylboronic acid, Pd(dppf)Cl2·CH2Cl2, Na2CO3, toluene, water, 90 °C; (c) NBS, AIBN, CCl4, hν; (d) NH4OH, ethanol, reflux; H2O; (e) Bu4NMnO4, pyridine, reflux.

The synthesis of 6-aza and 6-carbon analogues of 6-alkoxy-5aryl-3-pyridine carboxamides can be accomplished in only 3−4 steps starting from some of the above-described intermediates (Scheme 3). As an example, nucleophilic aromatic substitution of Scheme 3. Synthesis of Compounds 18, 21a



a

Reagents and conditions: (a) N-methyl-cyclopropanemethanamine, DBU, 90 °C then 2N NaOH; (b) B-4-chlorophenyl-boronic acid, (Ph3P)4Pd, Na2CO3, DME, water, 85 °C; (c) (1R,2R)-2-aminocyclohexanol hydrochloride, DMF, TBTU, DIEA, RT; (d) amine, TBTU, DIEA, DMF, rt; (e) (Ph3)2PdCl2, Cu(I)I, TPP-resin, DMF, MW, 120 °C; (f) Pd/C, H2, EtOAc.

RESULTS AND DISCUSSION All compounds were tested in triplicate in CB1 and CB2 receptor binding assays as previously described.25 Compounds with Ki values below 100 nM in CB1 receptor binding and lower than 20% inhibition at 3 μM at the CB2 receptor were further tested in a GTPγ[35S] assay in HEK cells overexpressing the human CB1 receptor. All compounds behaved as functional antagonists in these assays. We used ClogP to rank compounds for lipophilicity because the log D for the higher lipophilicity compounds could not be easily determined. Rat microsomal clearance was determined as described in the Experimental Section. Compound 4 with its cyclopentyloxy substituent was a moderately potent CB1 receptor antagonist. Systematic variation of the alkoxy substituent (Table 1) revealed that substituents larger than cyclopentyloxy lead to a 5−10-fold loss of potency, as shown for 14a and 14b in comparison with 14c, while smaller substituents as exemplified by 14d−e and 14g−i lead to up to 30−70-fold improved potency compared to 14c. More detailed examination of the SAR revealed that the size was not the determining factor; rather substitution adjacent to the oxygen atom is detrimental. Consistent with this observation, the 2-

7 with N-methyl-cyclopropanemethanamine in the presence of DBU, followed by saponification of the intermediate ester, yielded 5-bromo-6-[(cyclopropylmethyl)methylamino]-3-pyridinecarboxylic acid in 73% yield, which was then converted to 17 by Suzuki−Miyaura arylation followed by amide coupling to furnish 18. The 6-carba analogue of 14g, compound 21, could be prepared from 6-chloro-5-(4-chlorophenyl)-3-pyridinecarboxylic acid (9a) by conversion to the amide 19, followed by Sonogashira coupling with ethynylcyclopropane to afford 20, and finally palladium-catalyzed hydrogenation to yield 21. To explore the effect of varying the nitrogen atom position in the core on CB1 receptor affinity, we pursued the synthesis of isosteric 2-pyridincarboxamides and also prepared the corresponding pyrazine and pyridazine carboxamides (Schemes 4 and 5). D

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 4. Synthesis Substituted 2-Pyridine Carboxamides 32a,b and 37a,ba

Reagents and conditions: (a) CH2O, NaHCO3, reflux; (b) I2, NaHCO3; (c) R2-X, NaH, DMF, 90 °C; (d) arylboronic acid, Pd(dppf)Cl2·CH2Cl2, Na2CO3, toluene, water, 90 °C; (e) Bu4NBr, Zn, AcOH, 50 °C; (f) Bu4NMnO4, pyridine, 80 °C; (g) (1R,2R)-2-amino-cyclohexanol hydrochloride, DMF, TBTU, DIEA, rt; (h) arylboronic acid, Pd(dppf)Cl2·CH2Cl2, toluene, DMF, Na2CO3, water, 90 °C; (i) R2-OH, base, DMSO, rt; (j) H2O2, AcOH, 70 °C; (k) N,N-dimethyl-carbamic chloride, TMSCN, Et3N, CH3CN; (l) AcCl, EtOH, 90 °C; (m) NaOH, THF, water, 80 °C.

a

propyloxy compound 14f is 20-fold less potent than the nbutyloxy compound 14e, while compounds with bigger substituents such as benzyloxy- 14j and the heterocyclylmethyloxy- compounds 14k−n are all potent CB1 receptor antagonists. The alkoxy oxygen atom itself does not contribute to potency as both the N-methyl nitrogen analogue 18 and the carbon analogue 21 have potency similar to 14g. Both 18 and 21 are, however, inferior to 14g with respect to ClogP and metabolic stability in rat microsomes due to increased susceptibility to either oxidative N-demethylation or benzylic hydroxylation of 18 and 21, respectively. Of the compounds in Table 1, 14g−j, 14l, and 14o were the most stable in rat microsomes. The increased metabolic stability of these compounds may be, in part, due to the reduced overall lipophilicity compared to 14a. Alternatively, improved metabolic stability may be due to a reduced rate of hydroxylation of the CH2-group in α-position to the alkoxy oxygen by interference of neighboring groups (i.e., cyclopropyl for 14g, CF3 for 14i, γheteroatoms for 14h, 14l, and 14o). Compound 14g serves as reference to highlight the SAR of aryl- (Table 2), amide- (Table 3), and core- variations (Table 4) because we have the most complete set of data for analogues of 14g.

Moving the 4-chloro-atom to the 2-position on the phenyl ring, as in 15b, led to a 17-fold loss in potency compared to 14g, which could be partially compensated for by reintroduction of the 4-chloro-atom as in 15a, which led to only 11-fold loss of potency compared to 14g. Results with compound 15e suggest that double substitution in the 2- and 5-position partially restores potency; the 5-position substituent compensating for the negative impact of the 2-position substituent. Polar 4-position substituents were detrimental for CB1-receptor affinity (15k− n), with the 4-methoxy group leading to about 6-fold loss of affinity compared to 14g. However, all halogens and pseudohalogens in the 4-position seem to be equally acceptable, leading to compounds with potent CB1 receptor affinity. Metabolic stability in rat microsomes for 15g (4-CF3) and 15h (4-OCF3) was excellent, which, for 15h, was also corroborated by its high and sustained plasma exposure in rat SDPK studies (Supporting Information Figure S2). These data may indicate that modifications that reduce the propensity for arylhydroxylations yield metabolically more stable molecules suitable for in vivo experiments. Alternatively, the two relatively lipophilic compounds, 15g and 15h, may be protected from metabolic turnover by high protein binding. In some cases, microsomes actually under predicted the metabolic stability of these E

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 5. Synthesis of Compounds 44 and 51 with Pyrazine and Pyridazine Scaffold, Respectivelya

Reagents and conditions: (a) arylboronic acid, (Ph3)4Pd, Na2CO3, THF/water, 90 °C; (b) isoamylnitrite, TMSBr, CH2Br2, rt; (c) cyclopropanemethanol, NaH, DMF, rt; (d) Pd(dppf)Cl2·CH2Cl2, CO, NEt3, MeOH, 70 bar, 120 °C; (e) LiOH, water, THF, rt; (f) (1R,2R)-2amino-cyclohexanol hydrochloride, DMF, TBTU, DIEA, rt; (g) cyclopropanemethanol, NaH, DMSO, rt; (h) n-BuLi, TMP, I2, THF, −75 °C; (i) arylboronic acid, (Ph3)4Pd, Na2CO3, THF/water, 90 °C; (j) Pd(dppf)Cl2·CH2Cl2, CO, NEt3, MeOH, 70 bar, 120 °C; (k) LiOH, water, THF, rt. a

Table 1. In Vitro Pharmacology, ClogP, and Microsomal Stability for Compounds 14a−o

14a 14b 14c 14d 14e 14f 14g 14h 14i 14j 14k 14l 14m 14n 14o 18 21

R2

hCB1-R Ki [nM]

hCB2-R Ki [nM]

ClogP

Cl mic rat @ 2 μM [μL min−1 kg−1]

cyclohexyl phenyl cyclopentyl cyclobutyl n-butyl 2-propyl cyclopropylmethyl methoxyethyl 2,2,2-trifluoroethyl benzyl (3-methyl-isoxazol-5-yl)methyl pyrimidin-2-ylmethyl pyridin-4-ylmethyl pyridin-3-ylmethyl pyridin-2-ylmethyl

3035 ± 394 1657 ± 625 374 ± 223 208 ± 130 43.8 ± 11.0 1074 ± 127 11.4 ± 4.5 5.1 ± 1.6 12.4 ± 6.7 40.2 ± 14.7 10.6 ± 4.8 51.6 ± 24.5 58.7 ± 49.2 23.5 ± 8.5 139 ± 93 51.1 ± 20.3 16.7 ± 3.4

>10000 >10000 >10000 >10000 >10000 >10000 >7000 >10000 >7000 >10000 >10000 >10000 >10000 >10000 >10000 >10000 >10000

5.8 5.1 5.3 4.7 5.4 4.6 4.8 3.7 4.6 5.2 3.4 2.7 3.7 3.7 3.7 5.3 5.3

662 353 210 101 64 98 23 20 151 12 118 374 44 1730 1503

The SAR exploration of the amide side chain revealed that the rimonabant-like side chain (piperidin-1yl) is inferior to the (1R,2R)-2-hydroxycyclohexyl side chain of 14g, as shown by the 50-fold loss in potency observed with 16a. The enantiomer and the diastereoisomers of 14g (cf. 16c−e) lose considerable potency for the CB1 receptor. Suitable alternatives leading to

compounds, while rat hepatocyte clearance seemed to be more predictive for in vivo behavior. For example, rat hepatocyte clearance for 14g was 3.3 μL min−1 (106 cells)−1 @ 10 μM, and in vivo clearance and Vss in rat at a dose of 1 mg/kg iv was 7.4 mL min−1 kg−1 and 1.8 L/kg, respectively. F

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 2. In Vitro Pharmacology, ClogP, and Microsomal Stability for Compounds 15a−n

15a 15b 15c 15d 15e 15f 15g 15h 15i 15j 15k 15l 15m 15n 14g

R1

hCB1-R Ki [nM]

hCB2-R Ki [nM]

ClogP

2,4-dichloro 2-chloro 3,4-dichloro 3-chloro 2-chloro-5-CF3 4-fluoro 4-CF3 4-CF3O 4-methoxy 4-cyano 4-methanesulfonylamino 4-sulfamoyl 4-carboxy 4-carbamoyl 4-Cl

122 ± 49 190 ± 150 305 ± 201 129 ± 60 67.7 ± 21.4 42.8 ± 26.0 18.5 ± 1.9 23.6 ± 11.4 66.4 ± 11.2 23.6 ± 12.7 1900 ± 820 >10000 >10000 >8000 11.4 ± 4.5

>5000 >10000 1260 ± 320 >5000 2080 ± 480 >10000 >10000 >10000 >10000 >10000 >10000 >10000 >10000 >10000 >7000

5.2 4.5 5.4 4.8 5.4 4.2 5.0 5.2 4.1 3.5 3.0 2.4 3.9 2.7 4.8

Cl mic rat @ 2 μM [μL min−1 kg−1]

59 37 9 0−3

64

Table 3. In Vitro Pharmacology, ClogP, and Microsomal Stability for Compounds 16a−e

16a 16b 16c 16d 16e

R3

hCB1-R Ki [nM]

hCB2-R Ki [nM]

ClogP

piperidin-1-yl (2R)-2-cyclopropyl-2-hydroxypropyl (1S,2S)-2-hydroxycyclohexyl (1R,2S)-2-hydroxycyclohexyl (1S,2R)-2-hydroxycyclohexyl

563 ± 242 34.3 ± 27.7 247 ± 52 186 ± 75 104 ± 48

>10000 >10000 2637 ± 342 >5000 >7000

5.2 4.5 4.8 4.8 4.8

a

32a 37aa 44 51 14g

core (A1, A2, A3)

hCB1-R Ki [nM]

hCB2-R Ki [nM]

C, N, C N, C, C N, C, N C, N, N C, C, N

>7000 1830 ± 1369 25.8 ± 13.5 >5000 11.4 ± 4.5

1526 ± 701 863 ± 546 801 ± 171 >3000 >7000

2.4 163 111

provides the additional steric bulk in this position that is necessary to obtain high affinity CB1 receptor ligands. Compound 14g and some of its core variants (32a,b, 37a,b, 44, and 51) allow us to rationalize the CB1 binding pharmacophore in this compound class. The energy-minimized structure of pyrazine 44 in Figure 4 served as a template to explore the CB1 pharmacophore. In the minimum energy conformer, the orientation of the amide bond is locked via a combination of reduced steric repulsion and the favorable electrostatic interaction between the amide NH H-bond donor with the aromatic nitrogen H-bond acceptor.26 The amide bond in 14g has no preference for either of the two possible planar orientations and, therefore, 14g is readily able to adopt the same conformation as 44. The loss in potency observed with the isostere 37a highlights the relevance of the pyridine nitrogen in both 44 and 14g to mediate an attractive interaction with the CB1 receptor. The dramatic decrease of binding activity of the isosteric 2pyridinecarboxamide 32a and pyridazine 51 compared to 14g can be rationalized by the different relative orientation of the phenyl and amide substituents. The change in vector arrangement with respect to the aryl group and the amide carbonyl is depicted by the orange arrows overlaid on 44 and 32a in Figure 4.

Table 4. Pharmacological Data for Core Variants

compd

Cl mic rat @ 2 μM [μL min−1 kg−1]

a

Values measured for 32b and 37b confirmed a drastic loss of affinity for the CB1 receptor for these isosters of 14i.

CB1 affinity similar to the (1R,2R)-2-hydroxycyclohexyl side chain are many other simple hydroxy-alkyl side chains (data not shown). However, care must be taken to avoid introducing metabolically labile sites. One useful example is compound 16b, with its tertiary hydroxy group where the cyclopropyl group G

dx.doi.org/10.1021/jm4010708 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 5. Rat PK Data of Selected Compounds Compared with Rimonabanta Ki hCB1-R [nM] EC50 GTPγ[35S] hCB1-R [nM] rat Cl [mL min−1 kg−1] rat Vss [L/kg] rat plasma Cmax [ng/mL] rat plasma Tmax [h] F [%] B/P ratio rat PPB [%] rat plasma Cunbound,max [ng/mL] Cunbound,max [nM]

rimonabant

14g

14h

5.9 ± 4.9 1.5 ± 0.2c 54b 19.5b 244g 1.5g ndb 3−5 99.7 0.83g 1.8g

11.4 ± 4.5 3.9 ± 3.3d 7.4e 1.8e 729h 1.9h 23h ∼0.7 >98.4